Post by Abysmal HorrorPost by Neelakantan KrishnaswamiHi,
I've written an article on how to use causal influence diagrams to
construct situations and mechanics that enable significant player
choice during play.
<http://www.20by20room.com/2004/01/practical_causa.html>
Thank you for posting this link; I found your article to be
interesting and thought-provoking.
Thanks for the response. I'll respond to your comments slightly out of
order.
Post by Abysmal HorrorAlso, though the diagram provides a potentially useful range of
intermediate outcomes, I'd also note that a "botch" or "partial
success/failure" mechanic might serve a similar purpose.
A margin-of-success mechanic will give you a wider range of of answers
to the question "What happened?" What a causal network offers is a
systematic way of inventing answers to the question "What choices can
I, the player, make?" These are different things.
It's entirely possible to want an interesting set of choices to make
even when the outcome is binary. For example, imagine playing a lawyer
arguing a court case. Even though your PC's client can only be
aquitted or convicted, there is interesting decision-making (and hence
role-playing) interest in the sorts of arguments and legal strategies
the character can try.
Post by Abysmal HorrorBut which binary situations are usefully "enlarged"? Some players
would find the details of the hyperdrive's malfunction to be a
distraction, and would prefer to abstract the repair into a binary
"collect the modifiers and make the roll" resolution, so they could
get on with the "real" adventure, their escape or capture.
The answer to the question, "Which situations are usefully enlarged?"
is "Exactly the situations in which the players would be interested in
making choices for their PCs." The goal is to enable highly engaging
play, and you can't get that either if the players don't care about
the decisions they are asked to make, or if they don't get to make
decisions about the things they do care about.
It's worth reiterating that I'm talking about what the players care
about, rather than what the characters care about. You can't get
player engagement either if they don't care about the decisions they
have to make, or if they don't get to make decisions about what they
do care about.
Post by Abysmal HorrorYour example of a starship engineer trying to fix a disabled
hyperdrive seems to work well for adding a useful and not overly
complex set of decision parameters to this specific situation, but
I'm finding it hard to generalize from the example. I don't see how
I would take the model of a mechanical problem in a space opera game
and apply it, for instance, to a social problem in a fantasy game.
The basic trick to drawing up a causal influence diagram is to start
thinking in terms of cause and effect. The recipe is as follows:
1. Pick some things that you think might be interesting as outcomes in
play. Make sure you have a couple of possible alternative outcomes; in
order for a player's decisions to be consequential, they must have
consequences.
2. Think of things that could affect the outcome. Write them down, and
for each category you come up with, come up with a handful of possible
things that fall into that category.
3. Now, figure out which categories of events are direct causes of
which other events. A direct cause of a situation is something that
directly affects the outcome of that situation, without any mediating
categories. An indirect cause affects some other situation which in
turn affects the situation. To draw a causal influence diagram, you
only need to draw the direct causes -- the indirect influences are
shown by following the arrows from one category to the next.
There are two useful heuristics for identifying the causal
arrows. First, causes must preceed effects, so you can't have
something cause an event that happens before it. Second, a direct
cause must somehow "touch" its effect -- there is no "action at a
distance". Also, it's often easiest to think of a couple of distinct
final outcome situations (such as artistic success and popularity),
and then work backwards, figuring out the direct causes of each of
those outcomes.
One of my current projects is a game in which the player characters
are a rock group, and I'm trying to use influence diagrams to make the
music mechanics interesting to play with.
--
Neel Krishnaswami
***@cs.cmu.edu